There are fewer events more likely to induce a sense of déjà vu among weathered UK technology observers than the launch of a new government digital strategy.
During 14 years of Conservative-led governments, there were four: in 2010, overseen by Cabinet Office minister Francis Maude and written by entrepreneur and digital advisor Martha Lane Fox; in 2015, erstwhile the Coalition government was replaced by a full Tory administration; in 2017, for no apparent reason; and most late in 2022, after a reorganisation of digital transportation teams the year before.
So, erstwhile the Labour government announced that its first digital strategy was to be unveiled at a London event hosted by… Martha Lane Fox – the phrase “here we go again” inevitably sprang to mind.
But what – if anything – was different this time?
Perhaps the clearest indication that maybe, just maybe, this would not be the same as all the past, variously failed or thwarted attempts to bring public services full into the digital age, was the fact it was announced by a secretary of state for technology.
Peter Kyle may not be the first to have that occupation title – that fell to Michelle Donelan, erstwhile the Department for Science, Innovation and Technology (DSIT) was established by then prime minister Rishi Sunak in 2023. And Kyle is not the first minister to front a fresh digital strategy – but he is the first with full secretary of state credentials to do so.
A flurry of announcements
A flurry of tech announcements had built anticipation for the launch. Last week, prime minister Keir Starmer unveiled his AI action plan, putting artificial intelligence at the heart of economical growth and public service delivery.
Earlier in launch week, DSIT revealed a study by consultants at Bain & Company that analysed the ills of digital government in large item and clarity – highlighting £45bn of lost productivity through old, inefficient or just plain rubbish IT systems, processes and procedures.
And the day before Kyle’s set-piece launch, hosted in the Whitechapel office of the Government Digital Service (GDS), DSIT added an expansion of GDS’s remit and a first look at the Gov.uk app and digital wallet, including a fresh digital driving licence that can be utilized for identity and age verification on a smartphone.
As guests gathered for Kyle’s speech, 1 difference consecutive distant was the level of interest from the mainstream and political media – Computer Weekly can attest how different it is for the likes of Sky News, the Independent, Press Association, The Times, the Financial Times and Politics Home to show up for a press conference about digital government. Even The Sun was there. Times have changed in technology.
When Kyle initially described the plan as aiming to “transform the relation between citizen and state” it didn’t bode well. That exact phrase was utilized by erstwhile Cabinet Office minister Ben Gummer to launch his 2017 strategy.
But the chance for Computer Weekly to put the large question to a secretary of state for technology was surely novel. So, according to Peter Kyle, what’s going to be different this time?
Learning from experience
Dressed in a T-shirt and jeans, joking about turning up to a gathering of tech founders dressed the same way alongside a suit-wearing Starmer, Kyle was keen to put clear air between the Labour digital strategy and what went before.
“Quite simply, I’ve been learning from those experiences. I’ve gone to [former GDS chief] Mike Bracken and sat with him and learned from him. We’ve had Martha on tap. That 2010 drive from GDS was inspirational. They did deliver changes to government – and then it just stopped, it petered out, the political drive behind it disappeared. That has changed,” he said.
“I learned in opposition to go back and realize what the barriers to transportation have been and where things have gone wrong. erstwhile we came into power, we had a programme of government, we had a full series of implementation plans ready to go. And quite a few the work you see present wasn’t part of the implementation plans, but erstwhile I got a sense of the scale of what the capacity was, what can be created, then we started at it.”
“I learned in opposition to go back and realize what the barriers to transportation have been and where things have gone wrong. erstwhile we came into power, we had a series of implementation plans ready to go”
Peter Kyle, secretary of state for technology
It’s surely actual that fewer of Kyle’s predecessors with ministerial oversight of digital government would have taken the time to perceive to techies first. The Tory minister who launched the 2022 strategy cheerfully jumped on stage, admitted she knew nothing about technology, and proceeded to insult Birmingham and Blackpool. It’s fair to say Kyle was different on that, too.
“I’ve been working the phone, due to the fact that this is what Keir demanded of me,” he continued.
“So, where there are barriers, I’ve been speaking [for example] to government lawyers due to the fact that there are legal issues. erstwhile it came to the digital driving licence, I was told it couldn’t happen in 5 years in 1 Parliament. But erstwhile you start working the phone, and you have very strong relationships around the Cabinet table that all formed in opposition and have carried on into government, we – as secretaries of state – are talking an awful lot about this.
“Keir has been absolutely steadfast in demanding openness to the possibilities that the digital centre of government is making available. I can’t do this to government departments. I can only do it in partnership. We just battered distant at all of those different things.”
Snubbed noses
Kyle’s reflection that he can’t force his strategy onto another departments is notable – 1 of the frequent failings of erstwhile plans came erstwhile large digital teams specified as HM gross & Customs (HMRC) and the Department for Work and Pensions (DWP) fundamentally snubbed their noses at GDS and did their own thing. Whether deliberate or not, the presence of DWP secretary of state Liz Kendall in the front row of the launch event sent a strong signal that those old enmities will no longer be tolerated.
Kyle highlighted how Kendall sees a “jobcentre in people’s pockets” as 1 of the opportunities of the Gov.uk app. And for Kyle, that app in itself is simply a sign of how things have changed.
“The another large difference, which I’m certain people have noticed, is I’ve pre-announced a full set of very circumstantial products and services – government has never done that before,” he said.
“It [typically] says in general terms, we’re going to transform, we’re going to deliver – it doesn’t actually give a preview of what the product is going to look like and give a date for which it will land. Again, learning from the tech sector so that we reflect and respond.
“Why is it we [used to have] all these amazing things that are in train and then either it kind of drips out, or it gets out in the incorrect place or we wait until everything’s so battened down and over-tested before it’s released and nobody notices. Why can’t we just be honest with the public about what we’re doing behind the scenes? I thought everyone would be going, ‘No, no, no, don’t do it’. Nobody held me back from doing this.”
Not everybody is happy with this approach. Computer Weekly has heard from leaders in the digital identity sector, many of whom are furious that the government has announced a product for age verification on a mobile telephone through the digital driving licence. Kyle demonstrated prototypes of utilizing this for online buying – functionality that many firms have been encouraged by DSIT to make at large cost, and to certify against new government standards for digital IDs.
When asked about this possible conflict, a government authoritative read out a pre-prepared message – suggesting DSIT was aware of the likely reaction from the private sector – which fundamentally said, “We’re exploring all the options”. More controversy could yet emerge.
Core problems
But the success or failure of past digital government strategies has seldom been down to the technologies in place, or the political intent, or the amount of money available. The core problems have come in the way Whitehall works – or, more specifically, the way policymaking is conducted and the way technology budgets are spent.
Critics have long pointed out the deficiency of digital expertise active erstwhile policies are being discussed, and how tech teams typically end up presented with a fait accompli – here’s the policy, can you build us any digital stuff?
As digital government experts Jerry Fishenden and Cassian Young wrote in a fresh article for Computer Weekly, the process of deciding policy runs completely counter to the ways that digital approaches can improve those policies.
“Government’s linear approach to policymaking inevitably locks in questionable assumptions and constraints long before policy always makes contact with the real world. It’s an approach that generates immense missed opportunities – generalist politicians and officials frequently have small thought of how technology could make alternate ways of designing and delivering better policy outcomes,” they wrote.
Here, DSIT officials have been looking at global best practices to see how another governments have dealt with akin challenges, and are taking peculiar interest in Denmark’s approach of “digital-ready legislation”.
In Denmark, policymakers are told to admit “a request for increased awareness of whether the government can be managed digitally and on whether the implementation impacts have been clarified sufficiently”. In effect, it requires civilian servants to tell politicians erstwhile their desired policies could end up costing more money or introduce greater hazard if they do not consider the digital implications up front.
DSIT officials cited DWP and the Ministry of Justice as departments keen to pilot a fresh approach based on learning from Denmark’s example.
Non-agile spending
The way government spending and budgeting works is another area anathema to the agile approach that digital enables and requires. HM Treasury expects projects to have a fixed cost and timescale up front, with agreed benefits and timescales. Spending is classified as either “resource expenditure” for ongoing costs, or “capital expenditure” for investments. In a digital planet of prototypes and “test and learn”, of beta versions and continuous improvements, it’s a round, digital peg being forced into a square, bureaucratic hole.
Both these issues were highlighted in the survey commissioned by DSIT from consultants at Bain & Company into the state of digital government. The report is simply a thorough, open and honest read that acknowledges the failings that have led to multiple incomplete or failed strategies over the years.
“Funding models do not reflect modern digital practice,” it said. “Existing governance and controls processes are not seen as well suited to digital programmes as they prioritise predictable returns over longer-term resilience and continuous improvement, and do not let for the flexible nature of digital delivery.”
Feryal Clark, minister for AI and digital government in DSIT, told Computer Weekly this is going to change.
“We’re changing how we procure and make the case for digital investment due to the fact that presently you spend 18 months to 2 years building a business case to convince the Treasury to be able to invest, and by the time you go out to market, what’s in there is besides old and you can’t change it due to all the checks and balances,” she said.
“The way we procure is not agile, it’s not quick. And so we are completely overhauling that as well.”
Ministerial support
DSIT announced that the Treasury will “experiment” with a fresh approach that reflects the realities of digital delivery. DSIT officials said the Treasury has agreed to “small-scale” trials of fresh backing models in “a fewer departments”. Anyone with past experience of persuading the Treasury to effort new, agile spending models might fishy specified cautious language reflects the fact that Treasury mandarins stay little than keen.
Nonetheless, Kyle has another ministerial supporter in chief secretary to the Treasury Darren Jones – a erstwhile technology lawyer who worked for BT, and known to be a large digital advocate.
Jones, Kyle and Pat McFadden – minister for intergovernmental relations and seen as Starmer’s “fixer” and right-hand man – form a fresh “interministerial group” that oversees governance for digital services. erstwhile Labour came into power, this group issued a moratorium on all major government spending projects worth over £1m in the digital area, to “get a grip” – in Kyle’s words – on digital transformation across government.
McFadden has already announced a £100m fund to deploy “test and learn” approaches across the country. Jones, it appears, will be liable for getting a grip on the Treasury’s objections.
Generative AI
Another historical barrier to digital advancement has been data sharing, and here, too, Labour envisages changes, promising to mandate the publication of application programming interfaces (APIs) for digital systems – ironically, an thought first mooted by Lane Fox in her 2010 review.
Exposing APIs will besides aid another plank of the fresh DSIT approach – the usage of generative AI (GenAI) to support the work of civilian servants. As part of DSIT’s reorganisation of digital delivery, GDS has absorbed what was Number 10’s AI incubator team, a group of young and eager AI experts primarily recruited from outside the civilian service to identify opportunities where GenAI could improve productivity.
The AI squad has so far identified 57 possible applications for GenAI, of which 11 have reached prototype stage, with respective being put into active usage as part of a set of tools labelled “Humphrey” after the character from the BBC comedy Yes, Minister.
Online wags have already asked whether this means the GenAI apps will be supercilious, arrogant and think they know better than everyone, in line with the fictional Sir Humphrey Appleby they are named after. But the developers of the apps are far more evangelical. Since their decision into DSIT, they have already been asked to focus on how GenAI can improve citizen experience as well as cutting costs within the civilian service.
GenAI tools being introduced focus on summarising large bodies of information, specified as public consultations, Hansard records of Parliamentary debates, and ministerial meetings – the kind of “low-hanging fruit” that private sector digital leaders would say are a sensible place to start.
There are changes outside of Whitehall too. In the past, central government has mostly declined to work with local authorities on digital transformation – erstwhile iterations of GDS were expressly told not to engage with local government. The fresh plans affect reaching out to councils, to the degree that the Mayor of London’s chief digital officer, Theo Blackwell, has been seconded into GDS to work on engagement with councils.
A £45bn prize
The prize on offer for Kyle and the Labour government is the £45bn estimated savings that could be made through increased productivity if they get their digital strategy right. It’s clear that Kyle knows things not only request to be different, but request to be seen to be different.
“I could not have done any of the announcements I did present if digital services were constructed in the way they were erstwhile I came in as secretary of state,” he said.
“It is just a better way of doing government and that is part of the change that people have experienced and should be now noticing the difference of, in the change of government.”
So, is he right? Is this time different?
Labour has taken the time to learn the lessons of the past. Its plans hit most of the buttons that critics would have hoped to be addressed. The intent, the enthusiasm, the item of the strategy suggests that things are changing for the better.
But only erstwhile – if – this digital government strategy genuinely delivers on its promises, can it be said to be truly different.
